RDA-L Michael Gorman’s new essay on RDA

Posting to RDA-L

On 10/23/2015 3:31 PM, Brenndorfer, Thomas wrote:
The separation of data from display is a central idea. I think everyone should have had enough time by now to be clear-eyed about what the development of RDA was about. It is a bridge that supports the data that went into catalogs such that different implementation scenarios, including the original flat file card catalog design, can be realized.

The following article explains this point very well…


But the data has already been separate from the display since the implementation of MARC. I have worked with CDS-ISIS catalogs, which are similar to those early computerized catalogs because they are not a DBMS and they contain only native ISO-2709 records. Even in those databases, I could search and display any part of any record, or any groups of records, in any way I wanted. In MySQL databases, I have done the same thing, and with XML and other formats, I can do the same thing. In this sense, nothing is new since the 1960s.

As an example, I can write instructions to do anything, to put 100$d next to 300$b next to the entire 240 and end it all with a period or semi-colon or asterisk. Often I did this kind of thing to create .csv files (comma separated value) that I could then import into Excel or Access, but I could have done anything else with this information that I wanted. Once the bits of information in our catalog records were formatted separately with MARC format, everything could be searched and displayed however we wanted. And that happened in the 1960s.

Originally, in the US the purpose was to print cards but that was before video displays appeared. It didn’t have to just print cards however, it could have been displayed in lots of other ways. Where I worked with the CDS-ISIS database, they never printed cards but they used their ISO-2709 records to print out book catalogs. I am sure you could do exactly the same with MARC-21 records if you wanted to, and I think I may have done it myself at some point.

What any of this has to do with the rule changes mandated by RDA: the rule of one instead of rule of three, no more abbreviations, 245$b optional, adding relators, extra 33x fields etc., is beyond me. Except for the additional information mandated by RDA (33x fields, relators, etc. which are only on the newest records and therefore makes access inconsistent and unreliable) RDA won’t allow us to search or display anything that we haven’t been able to search and display for the last 50 or so years.

If we are talking about FRBR-type data structures, where the work is separate from the expression, etc. and everything is imported through URIs into a display that brings the WEMI together in some sort of fashion (I deliberately set aside issues of how genuinely useful that is for either searchers or librarians), it is still discussing differences in computer formats, not changes of the rules for describing resources.

If we are talking about linked data (heavenly chorus) then what we need to do is just … put in the links!

But RDA is not needed for any of that.